Low dose heparin — Low dose subcutaneous heparin for perioperative prophylaxis of venous thrombosis is usually given in a dose of 5,000 units two hours preoperatively and then every 8 or 12 hours postoperatively. An early prospective randomized study of over 4000 patients found that low dose heparin reduced the incidence of fatal pulmonary embolism in patients undergoing major surgical procedures from 0.7 to 0.1 percent compared to controls [25]. Pooled data from meta-analyses then confirmed that low dose heparin reduces the incidence of all deep vein thrombosis, proximal deep vein thrombosis, and all pulmonary emboli, including fatal pulmonary emboli [5,6,8]. Most of the patients in these trials underwent abdominothoracic surgery (particularly for gastrointestinal disease), but patients having gynecologic and urologic surgery, as well as mastectomy or vascular procedures, were also included.

Low dose heparin is also effective in reducing the incidence of venous thrombosis in nonsurgical patients. Two studies have shown that low molecular weight heparin was superior to placebo in the prevention of VTE in hospitalized medical patients with congestive heart failure, chronic obstructive pulmonary disease, sepsis, and a variety of other conditions who were immobilized for at least three days [26,27].

The incidence of major bleeding complications is not increased by low dose heparin, but there is an increase in minor wound hematomas. The platelet count should be monitored regularly in all patients on low dose heparin to detect the rare but significant development of heparin-induced thrombocytopenia.

In addition to the relatively low side effect profile, low dose heparin has the advantage that it is relatively inexpensive and easily administered. Anticoagulant monitoring is not required.

LMW heparin — A number of low molecular weight heparin preparations are available. (See "Clinical use of heparin and low molecular weight heparin"). These drugs have the advantage that they can be given once or twice a day at a constant dose without any laboratory monitoring. In addition, there is a lower incidence of thrombocytopenia than with unfractionated heparin. As an example, one randomized double-blind study of patients after hip surgery found that thrombocytopenia occurred in 9 of 332 patients (2.7 percent) receiving unfractionated heparin compared to none of 333 receiving low molecular weight heparin [28].

Low molecular weight heparin fractions have been evaluated for prophylaxis of venous thrombosis in a number of situations, primarily surgical:

Randomized clinical trials comparing low molecular weight heparin with unfractionated heparin in general surgical patients have found that low molecular weight heparins given once or twice daily are as effective or more effective in preventing thrombosis [7,8,29]. The incidence of major bleeding was similar for both classes of drug in most of the reports, although one study found an advantage for low molecular weight heparin when all bleeding end points were taken into consideration [29].
A number of randomized trials have compared low molecular weight heparin to unfractionated heparin, warfarin, acenocoumarol, or fondaparinux for the prevention of venous thrombosis following total hip replacement [7,8,16-22,29-35]. In North America, LMWH or warfarin are most commonly used prophylaxis for high risk procedures (eg, total joint replacement). In Europe, prophylaxis has routinely started preoperatively, while in North America, prophylaxis has been started postoperatively because of a concern for perioperative bleeding.
In the meta-analysis comparing preoperative with postoperative initiation of prophylaxis of DVT following total hip replacement surgery, it was shown that total DVT rates (but not proximal DVT rates) and major bleeding occurred significantly less frequently in the preoperative group compared with those who received postoperative prophylaxis [36]. LMWH started immediately before or early after surgery in patients undergoing total hip replacement resulted in significantly lower rates of both total and proximal DVT when compared with warfarin [35]. There was no difference between the preoperative and postoperative LMWH arms with respect to efficacy, but there was more major bleeding in the preoperative LMWH group when compared with warfarin.

Comparison across trials of patients with hip surgery has been difficult, since the drugs under investigation and their dosing schedules vary from one clinical trial to another. Even within the same clinical trial there can be considerable inter-center variability [30]. Bleeding rates vary quite widely across studies and different definitions are frequently used. The commonly used traditional classification for major or minor bleeding is recommended [30,31,36].

Although the number of patients undergoing total knee replacement now equals the number undergoing total hip replacement, there have been fewer trials in patients undergoing knee replacement [21,30,37,38]. Fondaparinux was shown to be superior to enoxaparin in reduction of total venous thromboembolism following total knee replacement surgery [39].
LMWH has been compared with compression stockings for the prevention of venous thromboembolism following neurosurgery [40,41]. In one study, LMWH plus compression stockings was superior to compression stockings alone [40] and in the other study LMWH was superior to the use of compression stockings [41].
Low molecular weight heparin significantly decreases the rate of total (31 versus 44 percent) and proximal (6 versus 15 percent) deep vein thrombosis when compared with unfractionated heparin in patients suffering multiple trauma [42].
Low molecular weight heparin (enoxaparin) appears as effective as subcutaneous unfractionated heparin as prophylaxis in elderly patients who are bedridden because of acute medical illnesses [43].
Low molecular weight heparin was shown to be as effective and safe as low dose heparin given three times a day in patients following acute ischemic stroke [44].
Oral anticoagulation — Oral anticoagulation (with warfarin) can be commenced preoperatively, at the time of surgery, or in the early postoperative period for prophylaxis of deep vein thrombosis. (See "Clinical use of warfarin") However, therapy started at the time of surgery or in the early postoperative period may not prevent small venous thrombi from forming during or soon after surgery, because the anticoagulant effect is not achieved until the third or fourth postoperative day [45]. Nonetheless, warfarin appears to effectively inhibit extension of such thrombi, if present, thereby preventing clinically important venous thromboembolism. The two step warfarin protocol appears to be no more effective than warfarin started the night before surgery in patients undergoing total knee replacement [46].

Warfarin has been compared with low molecular weight heparin in patients undergoing total hip replacement surgery [30-32,45-47]. Most studies showed superior benefit with LMWH, with two of the studies having a statistically significant difference favoring LMWH [32,46].
In studies comparing the efficacy and safety of LMWH with warfarin in patients undergoing total knee replacement the incidence of total deep vein thrombosis has been less in the LMWH group but the incidence of proximal venous thrombosis has been similar [29,30,38].

In a comparison of warfarin to external pneumatic compression after total hip replacement, warfarin was significantly more effective [48].
In a study of patients with hip fractures, warfarin was superior to aspirin or placebo for the prevention of deep vein thrombosis [49]. Similar results were noted in a post-discharge retrospective analysis of patients undergoing total hip arthroplasty; the administration of warfarin was associated with a lesser incidence of symptomatic venous thrombosis requiring rehospitalization compared to patients not receiving warfarin [50].
Intermittent leg compression — Intermittent pneumatic leg compression prevents venous thrombosis by enhancing blood flow in the deep veins of the legs, thereby preventing venous stasis. Pneumatic compression also reduces plasminogen activator inhibitor-1 (PAI-1) levels via an unknown mechanism and consequently increases endogenous fibrinolytic activity [51]. Thus, intermittent leg compression has both local and systemic effects. (See "Vascular endothelial function and the mechanisms of thrombolysis").